Advertising provides immense value to users by helping them find information and services quickly while enabling many free online services. Thus, the scope, importance, and reach of online advertising continues to grow at a rapid pace. On the flip side, this also attracts plenty of bad actors who misuse the ecosystem for deceptive purposes ranging from malware downloads, phishing attacks, tech scams, counterfeit goods, adult content, scareware popups and many more – by pushing bad ads in front of users. Attacks via ads are becoming more commonplace, and this is an area that needs continuous vigilance.
The user safety team at Bing Ads has people, processes, and automation behind the scenes to ensure ads on the Bing network are free from malicious content, to provide a safe browsing experience. Our policy team is spread over multiple continents to bring top notch policy compliance expertise for various markets. We also partner with other forensic teams within Microsoft for concerted action online and offline – allowing us to deter bad actors in the real world who often come back repeatedly in different forms.
Almost daily we all hear about working with big data and machine learning. Managing the quality of our ads means we are constantly pushing boundaries of both these technologies. These methods are a perfect fit to help us spot malicious intent within the millions of ads we receive every day. To put it in perspective, if one person took a minute to find and take down a bad ad or actor, it would take them nearly 500 years to remove the same number of bad ads or actors found by our automated methods in 2016.
Our machine learning automation scans and acts on all ads and advertisers that come to the platform – with occasional manual checks (since humans are still sometimes better in making certain complex judgements).
It is natural to ask – what is it about online advertising that requires these additional efforts? Why doesn’t advertising on TV or newspapers pose risk to users, at least to same extent? One reason is the anonymity provided by the internet. Online services can reach more users, at the same time anyone with internet access anywhere in the world can pretend to be an advertiser. Because of this, we put in additional effort to distinguish genuine advertisers from those who are not.
During 2016 we rejected over a hundred million ads for direct policy compliance issues or intention to mislead users. We introduced new policy around software download advertising that reduced unwanted and potentially malicious ads for many top free software programs. We ramped up systems that detect browser hijacking ads, phishing attempts, scareware ads, ads targeting the most common sites on the internet, and ads with multimedia content. We also enforced policies directed towards gender determination ads to comply with country specific regulations. As we start 2017, we have plans to evolve our methods to address new and emerging challenges.
Below is a summary of some of our enforcement actions in 2016
At Bing Ads we always value user feedback. You can report ads which you believe to be malicious or in violation of our policies through the feedback portal
As we make it harder for bad actors to show their ads, they will evolve to find new ways to continue their nefarious schemes. We are committed to remaining vigilant to ensure that our users are safe, and to deliver the best outcome for our valued advertisers across the Bing Ads marketplace.